HAL Id: hal-02024302
https://hal.uca.fr/hal-02024302
Submitted on 26 May 2021
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
3D electrical imaging of the inner structure of a complex
lava dome, Puy de Dôme volcano (French Massif
Central, France)
Angelie Portal, Y. Fargier, P. Labazuy, J.-F. Lénat, P. Boivin, D. Miallier
To cite this version:
Angelie Portal, Y. Fargier, P. Labazuy, J.-F. Lénat, P. Boivin, et al.. 3D electrical imaging of the inner structure of a complex lava dome, Puy de Dôme volcano (French Massif Central, France). Journal of Volcanology and Geothermal Research, Elsevier, 2019, 373, pp. 97-107. �10.1016/j.jvolgeores.2019.01.019�. �hal-02024302�
1
3D electrical imaging of the inner structure of a complex
1
lava dome, Puy de Dôme volcano (French Massif Central,
2
France)
3
A. Portal
1,4*, Y. Fargier
2,P. Labazuy
1, J.-F. Lénat
1,P. Boivin
1, D.
4
Miallier
35 6
[1] {Université Clermont Auvergne, CNRS, IRD, OPGC, LMV, F-63000 Clermont-Ferrand,
7
France}
8
[2] {GERS, IFSTTAR, Bron, France}
9
[3] {Université Clermont Auvergne, CNRS–IN2P3, LPC, F-63000 Clermont-Ferrand, France}
10
[4] {BRGM, DRP/IGT, 3 avenue Claude Guillemin, BP 36009, 45060 Orléans, France}
11
* Corresponding author at: BRGM, 3 avenue Claude Guillemin, 45060 Orléans Cedex 2, France. 12
Tel.: +33 (0) 2 38 64 32 34 13
E-mail address: a.portal@brgm.fr 14
Key words: Lava dome; Electrical Resistivity Tomography; 3D inversion; Puy de Dôme
15
volcano
16
Abstract
17
Lava domes result from extrusion of massive lava, frequent explosions and
18
collapses. This contribution focuses on a complex trachytic lava dome, the Puy de
19
Dôme volcano, located in the Chaîne des Puys volcanic field (French Massif Central,
20
France). We performed Electrical Resistivity Tomography (ERT) acquisitions on the
21
entire edifice in order to investigate its overall inner structure as well as to detail its
22
summit area. The resulting large ERT dataset integrated a recently developed 3D
23
inversion code based on an unstructured discretization of the geometrical model. The
24
3D inversion models obtained refine the existing geological model of the Puy de
25
Dôme's inner structure obtained by previous geophysical studies. These results also
26
highlight the strong fracturing and fumarolic alteration that affect the summit part of
27
the volcano.
28
1. Introduction
29
Volcanic lava domes are complex structures built up by highly viscous magmas
30
and formed by both intrusion and extrusion processes. During their growth,
31
gravitational instabilities create talus formed by rockfalls, and large collapses may
32
© 2019 published by Elsevier. This manuscript is made available under the Elsevier user license https://www.elsevier.com/open-access/userlicense/1.0/
Version of Record: https://www.sciencedirect.com/science/article/pii/S0377027318303561 Manuscript_0a2cb52ed6ce359397637a59169c3e19
2 trigger explosive eruptions (e.g. Mount St Helens 1980, Christiansen and Peterson,
33
1981; Soufrière Hills, Herd et al., 2005) and pyroclastic flows (e.g. Unzen volcano
34
1991, Sato et al., 1992). Because their construction is often incremental and/or
35
polyphase, lava domes are usually compound edifices. Even in the case of
36
composite lava domes whose construction has been monitored, their inner structure
37
remains difficult to establish because of the intercalation of massive lava, talus
38
breccia and pyroclastites, and also because endogenous processes, such as magma
39
intrusion or hydrothermal activity, cannot be observed at the surface. Nevertheless,
40
an understanding of volcanic dome construction and evolution is an important issue
41
for hazard assessment.
42
Here, we have used the Electrical Resistivity Tomography – ERT - method to
43
study the internal structure of a large Holocene lava dome, the Puy de Dôme, in the
44
French Massif Central. The ERT technique, initially developed for environmental
45
investigations and engineering (Chambers et al., 2006; e.g. Dahlin, 1996; Loke et al.,
46
2013 and references therein), is now widely used in volcanology (e.g.
Barde-47
Cabusson et al., 2014; Brothelande et al., 2015; Byrdina et al., 2018; Fikos et al.,
48
2012; Gresse et al., 2017; Soueid Ahmed et al., 2018). As a large range of resistivity
49
values is expected in lava domes, this imaging technique is well suited to the study of
50
their inner structure, as shown by the example of La Soufrière de Guadeloupe lava
51
dome (Brothelande et al., 2014; Lesparre et al., 2014; Nicollin et al., 2006).
52
This study presents the main results from ERT surveys performed on the Puy
53
de Dôme volcano between 2011 and 2014. Given the spatial geometry of the
54
datasets, we were able to carry out a 3D inversion approach, in order to better
55
constrain the inner structure of the volcano. For this purpose, we used a recent
56
inversion code developed by Fargier et al. (2017). Our inversion strategy was first to
57
study the whole lava dome, and then to focus on its summit area only. The geological
58
interpretation of the 3D inversion models provides new information about the dome's
59
inner structure. We propose a comparison between electrical resistivity models and
60
results obtained from gravity and magnetic measurements (Portal et al., 2016) and
61
discuss the synthetic geological model of the Puy de Dôme volcano.
62
2. Geological and structural settings
63
The Puy de Dôme volcano is located in the Chaîne des Puys volcanic field, the
64
most recent manifestation of the French Massif Central volcanism, composed of
3 around 80 aligned Quaternary monogenetic volcanoes (scoria cones, lava domes
66
and maars) (Boivin et al., 2017). The Puy de Dôme, an 11,000 years old trachytic
67
lava dome, is the largest edifice of the volcanic chain with an elevation of 1465 m, a
68
basal diameter between 1.5 and 2 km and an apparent height of 400 m (Fig. 1). The
69
dome is emplaced into a cluster of several scoria cones and their associated lava
70
flows (Boivin et al., 2017; Miallier et al., 2010; Portal et al., 2016). Geological studies,
71
based on field observations, initially propose a three-stage construction model for the
72
Puy de Dôme growth (Camus, 1975): 1) the growth of a first cumulo-dome, 2) the
73
partial destruction of its eastern part and 3) the growth of a second lava spine into the
74
resulting collapse scar. Recent works modify this model and suggest that the eastern
75
flank would result from a change in the eruptive dynamism and not from a flank
76
collapse (Boivin et al., 2017). Miallier et al. (2010) propose that the dome eruption
77
ended with a final explosive activity interpreted as a summit
78
phreatic/phreatomagmatic eruption. The rock alteration on several outcrops in the
79
summit area prove that a strong hydrothermal activity accompanied the Puy de
80
Dôme growth. Debris and/or pyroclastic flows also occurred as shown by the fans of
81
unconsolidated materials observed at the base of the volcano (Portal et al., 2016).
82
Finally, Boudon et al. (2015) suggest that the hydrothermal activity progressively led
83
to a silicified permeable lava dome through cristobalite deposition into the pores and
84
deep-seated fractures.
4 Fig. 1. Simplified map of the main morphological features observed on the Puy de Dôme volcano and its surrounding after the analysis of the high resolution DTM by Portal et al. (2016). The scoria cones limits are identified from Boivin et al. (2017). Coordinates: WGS84 – UTM31N.
The geological interpretation of the recent geophysical results (gravity and
86
magnetism) obtained by Portal et al. (2016) suggests that:
87
1. The upper part of the lava dome could be constituted of a carapace of solid
88
rocks, that is morphologically well-defined in the western part (Fig. 1);
89
2. The eastern flank of the volcanic dome, very regular from the top to the base
90
could be buttressed, in its summit part, by an underlying massive carapace
91
of rocks or welded pyroclastites;
92
3. The central part of the dome might be composed of successive massive
93
intrusions and extrusions of trachyte interbedded with collapse breccia.
94
3. ERT measurements
95
3.1. Data acquisition
96
We performed twelve ERT profiles of different lengths on the Puy de Dôme
5 volcano, between 2011 and 2014 (Table 1). The goal was to investigate the
98
geological structures at different scales (the entire edifice on the one hand and its
99
summit area on the second hand). We defined a unique central point at the top of
100
the volcano through which we connected all the profiles crossing the area. We used
101
a multi-electrode ABEM system (Terrameter SAS 4000) associated with an
102
electrode selector (ES10-64) for data acquisitions. A standard ERT configuration
103
was composed of 64 stainless steel electrodes. To improve ground/electrode
104
contact we used clay and salty water. We applied the alpha and
Wenner-105
Schlumberger protocols due to their good signal-to-noise ratio, their optimal depth
106
of investigation and their sensitivity to horizontal and vertical geological contrasts
107
(Dahlin et Zhou, 2004). We acquired every measurement at least three times, in
108
order to calculate a standard deviation on the data. A standard deviation value
109
greater than 5% (threshold fixed by the operator) led to additional measurements
110
(up to 4 stacks).
111
ERT lines were first deployed at the scale of whole dome (Fig. 2a). Two
112
perpendicular profiles allowed us to explore its inner structure (Table 1): an
113
approximately N-S 35 m electrode-spacing line (P1, 2.2 km-long) and a W-E line
114
(P2-1, 2.2 km-long), the latter performed in two stages. Indeed, an equipment
115
problem affected the measurements along the eastern part of the initial P2-1 profile
116
(beyond the 32th electrode). This led to data with very low signal-to-noise ratio that
117
we eliminated. To complete the truncated P2-1 dataset, we deployed a new line
118
from the summit to the volcano's eastern base (P2-2, 64 electrodes, 10 m electrode
119
spacing). We performed this 1.3 km-long line using two half-length roll-along. This
120
strategy results in a depth of investigation greater in the western part than in the
121
eastern one. Last, we performed two supplementary profiles on the southern flank
122
(P3) and at the eastern base (P4) of the volcano (Fig. 2 and Table 1).
6 Name Date Number of
electrodes
Electrode spacing
(m)
Coordinates (m, WGS84 - UTM31N) Start electrode End electrode
X Y X Y W h o le d o m e P1 06/2011 64 35 497466.54 5069700.19 496750.96 5067823.45 P2-1 06/2013 32 35 496175.36 5069129.13 497075.61 5068754.54 P2-2* 04/2014 128 10 497094.13 5068743.52 498169.52 5068440.54 P3 04/2014 64 10 497094.77 5068739.40 497254.84 5068236.55 P4 04/2014 64 10 497804.31 5068989.46 498043.67 5068428.31 S u m m it a re a P5* 06/2011 128 5 497177.05 5069013.00 496953.86 5068473.31 P6* 06/2011 128 5 496839.90 5068887.82 497380.95 5068691.14 P7● 01/2014 51 10 496962.63 5068997.07 497183.99 5068589.63 P8● 01/2014 64 5 497018.52 5068873.90 497173.54 5068606.14 P9° 04/2014 64 10 496855.39 5068562.44 497325.74 5068919.28 P10° 01/2014 64 5 496974.28 5068649.88 497214.65 5068845.12 P11 04/2014 64 5 497054.46 5068936.99 496939.20 5068652.20 *: roll-along acquisitions ; ●
and °: overlapped profiles
Table 1. Characteristics of the ERT profiles performed on the Puy de Dôme volcano, with variable electrode spacing.
We also carried out a detailed study of the summit area using profiles with
124
electrode spacing of 5 m (P5, P6, P8, P10 and P11) and 10 m (P7 and P9) (Table 1).
125
All the profiles intersecting at the same location as the long lines (Fig. 2b).
Half-126
length roll-along processes allowed us to extend two 5 m electrode spacing lines, P5
127
and P6. We could not expanded the P7 line beyond the 51th electrode because of the
128
presence of an access road and a railway.
129
We obtained the electrodes locations through differential GPS measurements
130
(GPS Topcon) with post-treatment of the data using the Topcon Tools software
131
(leading to centimetric precision in planimetry and altimetry). In some sectors, under
132
tree cover, we have extracted electrode elevation from the 0.5 m resolution DTM.
7
Fig. 2. (a) Location of the ERT profiles on the whole Puy de Dôme volcano. (b) Location of the ERT lines on the summit area. For P3 and P4 (a) we represent one electrode out of two. Coordinates: WGS84 – UTM31N.
8 3.2. Data processing and inversion
134
The reliability of the electrical resistivity distribution obtained from inversion
135
models significantly depends on the data quality. Efforts were made during field
136
measurements to lower data noise as much as possible by ensuring both good
137
electrode/ground contacts and setting robust acquisition parameters. Several
138
resistivity measurements were stacked and the resulting standard deviation q (also
139
called quality factor) is less than 1% for most of the datasets. Raw resistivity data
140
were filtered through a quality-based method (Brothelande et al., 2014), to eliminate
141
data characterized by a low signal to noise ratio. All the measurements with an
142
electrical potential difference of less than 1 mV and/or error higher than 5% were
143
rejected. Before inversion, we used X2IPI software (Robain and Bobachev, 2017) to
144
filter all datasets in order to remove artifacts due to the presence of strong
145
heterogeneities in the shallow levels of the measurements. Finally, visualization of
146
the data in pseudo-sections allowed us to eliminate the remaining spurious
147
measurements.
148
To perform the 3D inversion of our resistivity data, we use an inversion code
149
developed by Fargier et al. (2017). This algorithm is based on a conventional
Gauss-150
Newton smoothness-constrained method with an Occam-type regularization
151
(Constable et al., 1987; de Groot-Hedlin et Constable, 1990; Lines et Treitel, 1984). It
152
also uses a non-structured discretization method (tetrahedral mesh, Fig. 3; Rücker et
153
al., 2006), that is now widespread for inversion of ERT datasets, especially in
154
volcanology (e.g. Gresse et al., 2017; Revil et al., 2010; Rosas-Carbajal et al., 2016;
155
Soueid Ahmed et al., 2018). A complete description of this inversion code is given by
156
Fargier et al. (2017). The RMS (Root Mean Square) error, representing the difference
157
between the model response and the measured data, quantifies the reliability of the
158
inversion models.
159
9 Fig. 3. (a) Vertical section in the 3D mesh along the P1, north-south oriented profile (location of this section is indicated by the white arrow on b). (b) 3D surface mesh with topography of the Puy de Dôme volcano, the blue dots represent the location of the electrodes.
We divide the total data set (6709 measurements) in two to perform the 3D
161
inversion. The two derived datasets rely on the repartition and the geometry of the
162
acquisition profiles as well as on our knowledge on the complexity of the geological
163
structure of the dome. Thus, the first inversion named WDI (Whole Dome Inversion),
164
integrates the ERT profiles P1 to P4 (Table 1 and Fig. 2a). It aims to constrain the
165
overall structure of the entire edifice in order to highlight the main structures inside
166
the lava dome. The second inversion SAI (Summit Area Inversion) focuses on the
167
summit lines P5 to P11, (Table 1 and Fig. 2b) to detail the summit part of the
168
volcano. For each inversion set, we fix a homogeneous initial model for the first
169
iteration with a resistivity equal to the mean resistivity of the input dataset. Each
170
following iteration uses the model of the previous one as reference. Our approach to
171
treat the data independently (two distinct initial model) lies on the ambition to limit the
172
propagation of error from the first model to the other.
173
10
4. Results
175
The inversion model of the whole Puy de Dôme (WDI) is associated to a
176
global RMS error of 7.8%. We extract horizontal sections (Fig. 4) and vertical ones
177
(Fig. 7, Fig. 8a and Fig. 9a) in the 3D model. The global RMS error of the detailed
178
3D inversion model of the summit area (SAI) is 12.7%. Horizontal (Fig. 5) and vertical
179
(Fig. 6) sections of this model have been extracted for description.
180
The presence of many man-made structures (roads, rails, paths…) affects the
181
inversion models. We identified resulting artifacts (highly conductive patches, red
182
triangles on Fig. 6 to Fig. 9) that will not take part of the following
183
description/interpretation. Human activity has also strongly reworked the summit area
184
of the Puy de Dôme, as evidenced by archeological vestiges and buildings. This
185
support our choice to not describe and interpret any resistivity anomaly in the first ten
186
meters of the SAI models (Fig. 6). Although the inversion integrates the loss of
187
information and constraints between profiles and with depth by an increase of both
188
the tetrahedrons size (Fig. 3) and the smoothing factor, we delineate opacity masks
189
on horizontal (Fig. 4 and Fig. 5) and vertical sections (Fig. 6 and Fig. 7). We
190
delineate a buffer area along the profiles for the horizontal sections (the buffer
191
distance equal to twice the electrode spacing). For the vertical sections, we used the
192
data distribution with depth taking into account the topography. The objective of the
193
masks is to focus the description and interpretation of the models in the better
194
constrained areas.
11 Fig. 4. Horizontal sections of the 3D inversion model of the Puy de Dôme (WDI). Sections every 50 m, starting from the elevation of 1300 m (top left) to 1150 m (bottom right). Black dots indicate the position of the electrodes. Opaque zones delineate the areas less constrained by ERT measurements. Coordinates: WGS84 – UTM31N.
12 Fig. 5. Horizontal sections of the 3D inversion model of the summit area of the Puy de Dôme (SAI). Sections every 25 m, starting from the elevation of 1425 m (top left) to 1350 m (bottom right). Black dots indicate the position of the electrodes. Opaque masks delineate the areas less constrained by the ERT measurements. Coordinates: WGS84 – UTM31N.
13
Fig. 6.Vertical sections extracted from the 3D inversion model of the summit area of the Puy
de Dôme volcano (SAI): (a) P7-P8, (b) P5, (c) P6, (d) P9-P10, (e) P11. (f) Map of the location of the ERT lines. Opaque masks delineate the areas less constrained by the ERT measurements. Red triangles indicate local conductive patches associated to man-made structures (roads, paths, rail…). White lines refers to horizontal slices on Fig. 5. Rn and Cn represent specific features related to resistive and conductive bodies, respectively.
14 There is a wide range of resistivity variations at the scale of the Puy de Dôme,
196
from conductive values (ρ~100 Ohm.m) to resistive ones (ρ~10 kOhm.m). The mean
197
resistivity of the dome is about 2000-3000 Ohm.m.
198
We identify several highly resistive structures (ρ>5000 Ohm.m):
199
- The R1 and R2 bodies are present near the model edges (Fig. 4 and
200
Fig. 8a);
201
- A large highly resistive (ρ>9000 Ohm.m) body, R3, is identified in the lower
202
part of the western flank (Fig. 4 and Fig. 9a). This structure is visible around
203
an elevation of 1250 m and could extend beyond the maximum depth of
204
investigation;
205
- To the North, a thin and superficial (around 30 m thick, Fig. 4 and Fig. 8a)
206
R4 structure follows the slope of the volcano;
207
- To the South, we observe two larger highly resistive bodies. The first one,
208
R5, seems to be restricted to the flank (maximum thickness of 120 m). The
209
second, R6, reaches the surface (Fig. 5 and Fig. 6) and extends inside the
210
volcano (Fig. 7a and Fig. 8a). Between an elevation 1200 m and the
211
surface, those two units progressively connect (Fig. 4 and Fig. 5).
212
- A well-delimited resistive structure (R7) occupies the upper part of the
213
western flank of the lava dome (Fig. 4, Fig. 5, Fig. 6 and Fig. 9a). Another
214
resistive structure body, R8, develops along the upper part of the NW flank
215
(Fig. 5 and Fig. 6).
216
- Last, we highlight a very large resistive structure along and within the
217
eastern flank. We can distinguish three sub-units. The large and deep R9
218
(Fig. 4 and Fig. 8b) structure may extends beyond the maximum depth of
219
investigation (depth>235 m in the central part of this body). In the summit
220
area, the surface R9’ body is around 30 m thick (Fig. 6a) as well as the R9”
221
structure at the bottom of the edifice (Fig. 7b).
222
A low-resistivity pattern, C1 (ρ<1000 Ohm.m), is observed in the upper part of
223
the dome, beneath the summit area (Fig. 7a, Fig. 8a and Fig. 9a). The horizontal
224
slices of the detailed SAI model (Fig. 5) show that several highly resistive bodies
225
intersect this C1 structure.
15 The detailed summit 3D model (Fig. 5 and Fig. 6) maps more accurately the
227
main resistant structures identified on the entire dome (R4, R5, R6, R7, R8 and R9’).
228
The R7 resistive structure extends to the center part of the dome (Fig. 6c) and shows
229
an elongated shape in the W-E direction (Fig. 5). The R9’ unit also presents an
230
elongated shape (at an elevation of 1400 m, Fig. 5) from the dome’s center toward
231
the East.
232
Fig. 7. Vertical sections extracted from the 3D inversion models of the entire Puy de Dôme edifice (WDI). (a) P3 profile. (b) P4 profile. (c) Map of the location of the two ERT lines. Opaque masks delineate the areas less constrained by the ERT measurements. White lines (a) refers to horizontal slices on Fig. 4. Rn and Cn represent specific features related to resistive and conductive bodies, respectively.
5. Interpretation and discussion
16 We base the following interpretation on the resistivity models described above,
234
while comparing them to those of the previous geophysical results (gravity and
235
magnetism, Portal et al., 2016). We first focus on the volcanic formations identified
236
below the Puy de Dôme volcano. Then, we discuss the overall structure of the lava
237
dome before concentrating on the summit part of the Puy de Dôme.
238
5.1. The surrounding volcanic structures
239
Within the flanks, we identify high resistivity zones (ρ>5000 Ohm.m). According
240
to morphological analysis (Fig. 1), field observations (presence of red scoriae and
241
massive bombs) and previous geophysical results (low density body - 1.6, Fig. 8b
242
and Fig. 9b; Portal et al., 2016), we can unambiguously interpret R1 and R2 as
243
underlying scoria cones (the Puy Lacroix and the Petit Puy de Dôme respectively).
244
The R3 resistive anomaly, identified on the western flank, also coincides with a
low-245
density body (1.4, Portal et al., 2016). Electric and gravity results confirm the
246
observations made by Miallier et al. (2010) who identify a buried cinder cone in this
247
area, named Cône de Cornebœufs. Our results support and confirm the theory that
248
the Puy de Dôme has grown on top of an area previously occupied by a swarm of
249
cinder cones (Boivin et al., 2017; Portal et al., 2016).
250
We observe that the dimensions of the R1 to R3 anomalies are more limited
251
than the dimensions of the low-density structures identified by Portal et al. (2016)
252
(Fig. 8 and Fig. 9). That lets us suppose a resistivity gradient inside the low-density
253
structures, from the surface toward the lava dome’s core. Considering the decrease
254
of the model resolution at depth, we can also hypothesize about a geological origin of
255
this resistivity evolution. This could correspond to an alteration of the existing scoria
256
cones by hydrothermal fluids during the lava dome’s growth. Indeed, summit
257
outcrops show evidences of a former fumarolic activity (ochre alteration of the
258
trachyte). The resistivity gradient could also reflect the variation of the water content
259
in the porous scoria formations (piezometric level). However, complementary data
260
are necessary to constrain the numerical and/or the geological contribution of the
261
resistivity gradient identified in the low-density structures.
17 Fig. 8. (a) Vertical section extracted from the 3D inversion models of the entire Puy de Dôme edifice (Whole Dome Inversion – WDI) along the P1 ERT profile. (b) Comparison of the electrical results to the corresponding gravity and magnetic ones from (Portal et al., 2016). (c) Map of the location of the P1 line (blue) and the corresponding section extracted from gravity and magnetism results (cyan dotted line). Opaque masks delineate the areas less constrained by the ERT measurements. Red arrows indicate local conductive patches associated to man-made structures (roads, paths, rails…). White lines (a only) refer to horizontal slices on Fig. 4. Rn and Cn represent specific features related to resistive and conductive bodies, respectively.
18 5.2. New constrains on the overall geological structure of the lava dome
263
Overall, the resistivity structure of the Puy de Dôme itself highlights several
264
main features: (1) very high resistivity surface or shallow layers on the summit and
265
flanks, (2) an overall resistive interior and base of the edifice, and (3) a low-resistivity
266
zone in the upper part of the dome, beneath the summit area.
267
The high resistivity units R4 (its upper part, above and elevation of 1350 m,
268
Fig. 6b), R5, R6, R7, R8 (resistivity >5000 Ohm.m, Fig. 4, Fig. 6, Fig. 8a and
269
Fig. 9a) coincide with steeply sloping areas. Morphologically, these zones
270
correspond to surface massive trachyte ridges (Fig. 1), suspected to extend at
271
shallow depth. The mentioned resistive patterns correspond to low density structures
272
(1.4 to 1.6; Portal et al., 2016). The upper part of R4 and the R6 body also show a
273
remanent magnetization of about ~ 5 A/m (Fig. 9b). We suggest that these resistive
274
formations are composed of massive trachytic lava bodies. They could be former
275
lava intrusions emplaced during the construction of the spiny dome. At the scale of
276
the entire edifice, the electrical results do not highlight the presence of a massive
277
trachytic carapace as initially proposed by Portal et al. (2016).
278
The northern flank is globally less resistant (Fig. 8a) with no specific density or
279
magnetic signature (Fig. 8b). It also present a surface morphology smoother than
280
that of the southern and eastern flanks (Fig. 1). These observations suggest that the
281
northern flank would be composed of slightly different material probably with very few
282
or no massive lava and possibly more talus breccia. Below the elevation of 1350 m,
283
the shallow resistive pattern R4 (ρ>5000 Ohm.m) seems to be associated to an
284
intermediate density – 2.0 - structure and partially magnetized body (5 A/m, Fig. 8b,
285
Portal et al., 2016). This resistive formation would correspond to recent pyroclastic
286
density current deposit (Boivin et al., 2017; Portal et al., 2016). Boivin et al. (2017)
287
also describe the presence of tephra-fall deposits in this area, issued from the
288
Kilian’s crater eruption, and that could contribute the R4 resistive response.
289
Our results show that the eastern flank, whose morphology is singular (Fig. 1),
290
has a specific and complex high resistivity signature. Indeed, we identify the thick
291
layer R9, between 350 m and 850 m of distance along the profile (Fig. 9a) and the
292
thinner R9’ (Fig. 6c and Fig. 9a) and R9” (Fig. 7b) bodies. In morphology, the
293
eastern flank looks like a nearly perfect half cone with a mean slope of about 33-35°
19 (Boivin et al., 2017; Portal et al., 2016). Such a value is too high for a repose angle of
295
loose material (for comparison, the nearby cinder cones have average slopes of less
296
than 25°). Boivin et al. (2017) propose that this eastern flank is mostly composed of
297
consolidated cinder deposits originating from a second exogenous eruptive phase of
298
the dome’s construction. According to this hypothesis we suggest that the R9’, R9”
299
and the surface part of the R9 (first 30 m) correspond to this welded cinder
300
pyroclastite deposits. Following this eruptive scheme, the high slopes of the eastern
301
flank could be due to an immediate induration process of the cinder products (the
302
mechanisms of such a phenomenon are still under investigation, Boivin et al., 2017).
303
The rest of the R9 signature (below 30 m from the surface) could correspond to
304
unaltered breccia with massive lava intrusion contemporaneous to the cumulo-dome
305
construction. The lack of morphological evidences of spiky dome features along the
306
eastern flank could result from the pyroclastite emplacement that fill and cover the
307
trachyte ridges. Finally, deep inside the eastern flank of the dome, the density model
308
also suggests a low-density signature (Fig. 9b) interpreted as strombolian deposits
309
(Portal et al., 2016). The high resistivity observed in this area (lower part of R9,
310
Fig. 9a) support this interpretation without giving discriminating criterion.
311
The rest of the dome’s inside has globally relatively high resistivity values (from
312
about 2000 Ohm.m up to 5 000 Ohm.m). Portal et al. (2016) show that these parts of
313
the dome have a generally low density (around 1.8, Fig. 8b and Fig. 9b) except the
314
presence of a dense (2.1) and magnetized (5 A/m) core. Although the model is less
315
constrained at depth, it seems that the high resistivity values observed inside the lava
316
dome support their interpretation: a conduit composed mainly of massive, poorly
317
fractured and/or altered rocks surrounded by a cogenetic breccia. The corresponding
318
high resistivity values identified suggests that the breccia probably contains former
319
massive intrusions. While the resistivity signature does not allow differentiating both
320
the breccia (low density) and the conduit (high density and magnetization), the latter
321
seems characterized by resistivity values globally slightly lower than the containing
322
formations. The hypothesis proposed here is that, along the conduit, the resistivity
323
signature may result from rock alteration due to fluid circulations during the dome
324
growth and evolution.
325
Finally, the central upper part area of the dome is very different from the flanks.
326
The sections in Fig. 8a and Fig. 9a clearly show that this zone is the most conductive
20 part of the edifice (C1, 100-1500 Ohm.m). On the long ERT profiles (P1 and P2), the
328
outlines of C1 are well highlighted. However, it is with the detailed summit ERT
329
profiles that the complex geometry and organization of this zone can be deciphered
330
(Fig. 6).
331
Fig. 9. (a) Vertical section extracted from the 3D inversion models of the entire Puy de Dôme edifice (Whole Dome Inversion – WDI) along the P2 ERT profile. (b) Comparison of the electrical results to the corresponding gravity and magnetic ones from (Portal et al., 2016). (c) Map of the location of the P2 line (red) and the corresponding section extracted from gravity and magnetism results (yellow dotted line). Opaque masks delineate the areas less constrained by the ERT measurements. Red arrows indicate local conductive patches associated to man-made structures (roads, paths, rails…). White lines (a only) refer to
21 horizontal slices on Fig. 4. Rn and Cn represent specific features related to resistive and conductive bodies, respectively.
5.3. The complex summit area
332
To analyze the sections in Fig. 6, we have to keep in mind that the thin (a few
333
meters to a few tens of meters depth), highly resistive or conductive layers in this
334
zone are strongly affected by various man-made structures and reworking.
335
The conductive zone C1 occupies the central part of the summit area, but its
336
shape, as well as the local variations in resistivity, are complex (Fig. 5). However, it
337
exhibits clear characteristics:
338
- Considering its dimensions and its bulk resistivity, it constitutes a major
339
structure;
340
- It has a maximum vertical extent of about 200 m;
341
- Its peripheral vertical limits are sharp;
342
- It may be composed of many sub-units separated by resistive structures
343
(Fig. 5 and Fig. 6b, c and d).
344
Therefore, C1 could represent a single unit with resistive bodies embedded in it.
345
Field observations show evidences of small fissures and fumarolic alteration in the
346
upper part of the dome (Miallier et al., 2010; Portal et al., 2016). The hydrothermal
347
activity is commonly observed on recent or active lava domes and associated to a
348
conductive signature of the corresponding deposits (e.g. Bedrosian et al., 2007;
349
Byrdina et al., 2017; Rosas-Carbajal et al., 2016; Zlotnicki et al., 1998). We therefore
350
suggest that the C1 anomaly is evidence of the presence of a former hydrothermal
351
system in the upper part of the lava dome resulting into high fracturing combined to
352
an important fumarolic alteration of the rocks. Nevertheless, the comparison between
353
electrical models and results presented by Portal et al. (2016), show that this
354
conductive body C1 is not correlated to a specific density or magnetic pattern (Fig. 8
355
and Fig. 9). Instead, they show the presence of dense, highly magnetized rocks in
356
the central part of the dome, from the surface to possibly the base of the edifice.
357
Because the data coverage in both gravity and magnetic data is high in the summit
358
area, the shallowness of the top of the dense and magnetized bodies identified in the
359
models is reliable. Moreover, fracturing and hydrothermal fluids circulations usually
360
contribute to lower both the density and the magnetization of rocks (e.g. Bouligand et
22 al., 2014). To support resistivity, gravity and magnetic signatures, we propose the
362
following hypothesis. The fumarolic alteration is concentrated along a network of
363
small-cracks observed in the field as well as in the upper part of the deep-seated
364
fractures network identified by Boudon et al. (2015), surrounding unaltered rocks (still
365
dense and highly magnetized). In this case study, electrical results provide significant
366
arguments on the level of rock alteration in the upper part of the dome.
367
6. Conclusion
368
The ERT imaging of Puy de Dôme volcano aims at investigating the overall inner
369
structure of the dome as well as its summit area. The resulting datasets is large, and
370
such a density of measurement is rare regarding the study of lava domes. Here we
371
present the results of a 3D inversion of this electrical datasets as well as a
372
confrontation with complementary geophysical results from Portal et al. (2016).
373
Besides to confirm some elements of the synthetic model of the inner structure of the
374
Puy de Dôme volcano proposed by Portal et al. (2016), the geological interpretation
375
of the electrical results provide new details. The presence of massive units inside the
376
collapse breccia are evidences of former trachytic intrusions, typical of an
377
endogenous construction. Our results provide also precisions about the geometry of
378
the deposits that covered the eastern flank, which definitively excludes a major flank
379
collapse of the lava dome (Camus, 1975) and which allowed Boivin et al. (2017) to
380
propose the presence of a welded cinder deposits issued from a second exogenous
381
eruptive phase. Then, we highlight, for the first time, the boundaries of the former
382
hydrothermal system of the Puy de Dôme volcano, focused in its summit part. The
383
hydrothermal alteration also affects the feeding conduit of the lava dome, with an
384
alteration that decreases with depth.
385
More generally, this study greatly contributes to our knowledge about the
386
formation of volcanic domes although it appears difficult to draw a general model of
387
such a complex phenomenon. It seems that the magmatic feeding is concentrated
388
along an eruptive conduit. Its localization strongly depends of the volcano substratum
389
and can evolve under the pressure of the accumulated volcanic deposits. The later
390
constitute a substantial volume of the edifices as already observed during recent
391
eruptions (e.g. Soufrière Hills volcano; Wadge et al., 2009).
23 This resistivity study of a complex volcanic edifice, as well as the associated
393
gravity and magnetic study (Portal et al., 2016) are important for evaluating the
394
capacity of geophysical methods to explore the interior of a volcano, and to define
395
the best strategy to implement for that purpose. For each method, and particularly for
396
resistivity, the necessity to have good data coverage is essential to be able to derive
397
3D models and characterize structures at different scales. Even with good coverage,
398
models uncertainties can make the geological identification of structures difficult,
399
especially with increasing depth. We thus prove that collecting data, which are
400
sensitive to different physical parameters (resistivity, density and magnetization),
401
constitutes a powerful means for discriminating the geology of structures that would
402
otherwise be impossible to distinguish with one parameter (e.g. the density allows us
403
to differentiate resistive porous unsaturated rocks from scoria deposits). This case
404
study of the Puy de Dôme is therefore important at two levels: for providing
405
information about the architecture of a complex lava dome, and for guiding the
406
strategy for studying other volcanic edifices.
407
Acknowledgements
408
The LIDAR data used in this study derive from a collective project driven by the
409
Centre Régional Auvergnat de l’Information Géographique (CRAIG), financially
410
supported by the Conseil Départemental du Puy-de-Dôme, the European Regional
411
Development Fund and the University of Clermont-Ferrand. Datasets are available
412
on request to the authors. We thank the students and permanent staff of the
413
Laboratoire Magmas et Volcans (LMV), the TOMUVOL collaboration and the
414
Observatoire de Physique du Globe de Clermont-Ferrand (OPGC) for participation
415
and logistics during field surveys. Finally, we thank the anonymous reviewer for
416
helping comments to improve this manuscript. This research was supported by the
417
French Government Laboratory of Excellence initiative n°ANR-10-LABX-0006, the
418
Région Auvergne and the European Regional Development Fund. This is Laboratory
419
of Excellence ClerVolc contribution number 308.
24
References
421
Barde-Cabusson, S., Gottsmann, J., Martí, J., Bolós, X., Camacho, A.G., Geyer, A.,
422
Planagumà, L., Ronchin, E., Sánchez, A., 2014. Structural control of
423
monogenetic volcanism in the Garrotxa volcanic field (Northeastern Spain) from
424
gravity and self-potential measurements. Bull. Volcanol. 76, 1‑13.
425
https://doi.org/10.1007/s00445-013-0788-0
426
Bedrosian, P.A., Unsworth, M.J., Johnston, M.J.S., 2007. Hydrothermal circulation at
427
Mount St. Helens determined by self-potential measurements. J. Volcanol.
428
Geotherm. Res. 160, 137‑146. https://doi.org/10.1016/j.jvolgeores.2006.09.003
429
Boivin, P., Besson, J.-C., Briot, D., Deniel, C., Gourgaud, A., Labazuy, P., de
430
Larouzière, F.-D., Langlois, E., Livet, M., Médard, E., Merciecca, C., Mergoil, J.,
431
Miallier, D., Morel, J.-M., Thouret, J.-C., Vernet, G., 2017. Volcanology of the
432
Chaîne des Puys. Parc Nat. Régional la Chaîne des Puys (Ed.), Cart. Fasc. 6e
433
édition 200pp.
434
Boudon, G., Balcone-Boissard, H., Villemant, B., Morgan, D.J., 2015. What factors
435
control superficial lava dome explosivity? Sci. Rep. 5, 14551.
436
https://doi.org/10.1038/srep14551
437
Bouligand, C., Glen, J.M.G., Blakely, R.J., 2014. Distribution of buried hydrothermal
438
alteration deduced from high-resolution magnetic surveys in Yellowstone
439
National Park. J. Geophys. Res. Solid Earth 119, 2595‑2630.
440
https://doi.org/10.1002/2013JB010802
441
Brothelande, E., Finizola, A., Peltier, A., Delcher, E., Komorowski, J.-C., Di Gangi, F.,
442
Borgogno, G., Passarella, M., Trovato, C., Legendre, Y., 2014. Fluid circulation
443
pattern inside La Soufrière volcano (Guadeloupe) inferred from combined
444
electrical resistivity tomography, self-potential, soil temperature and diffuse
445
degassing measurements. J. Volcanol. Geotherm. Res. 288, 105‑122.
446
https://doi.org/10.1016/j.jvolgeores.2014.10.007
447
Brothelande, E., Lénat, J.-F., Normier, A., Bacri, C., Peltier, A., Paris, R., Kelfoun, K.,
448
Merle, O., Finizola, A., Garaebiti, E., 2015. Insights into the evolution of the
449
Yenkahe resurgent dome (Siwi caldera, Tanna Island, Vanuatu) inferred from
450
aerial high-resolution photogrammetry. J. Volcanol. Geotherm. Res. 299, 78.
451
https://doi.org/10.1016/j.jvolgeores.2015.04.006
452
Byrdina, S., Friedel, S., Vandemeulebrouck, J., Budi-Santoso, A., Suhari, Suryanto,
453
W., Rizal, M.H., Winata, E., Kusdaryanto, 2017. Geophysical image of the
454
hydrothermal system of Merapi volcano. J. Volcanol. Geotherm. Res. 329,
455
30‑40. https://doi.org/10.1016/J.JVOLGEORES.2016.11.011
456
Byrdina, S., Grandis, H., Sumintadireja, P., Caudron, C., Syahbana, D.K.,
457
Naffrechoux, E., Gunawan, H., Suantika, G., Vandemeulebrouck, J., 2018.
458
Structure of the acid hydrothermal system of Papandayan volcano, Indonesia,
459
investigated by geophysical methods. J. Volcanol. Geotherm. Res. 358, 77‑86.
460
https://doi.org/10.1016/J.JVOLGEORES.2018.06.008
461
Camus, G., 1975. La Chaîne des Puys - Étude structurale et volcanologique.
462
Université de Clermont.
463
Chambers, J.E., Kuras, O., Meldrum, P.I., Ogilvy, R.D., Hollands, J., 2006. Electrical
25 resistivity tomography applied to geologic, hydrogeologic, and engineering
465
investigations at a former waste-disposal site. Geophysics 71, B231 B239.
466
https://doi.org/10.1190/1.2360184
467
Christiansen, R.L., Peterson, D.W., 1981. Chronology of the 1980 eruptive activity.
468
US Geol. Surv. Prof. Pap 1250, 17 30.
469
Constable, S., Parker, R.L., 1987. Occam’s inversion: A practial algorithm for
470
generating smooth models from electromagnetic sounding data. Geophysics 52,
471
289 300.
472
Constable, S.C., Parker, R.L., Constable, C.G., 1987. Occam’s inversion: A practical
473
algorithm for generating smooth models from electromagnetic sounding data.
474
Geophysics 52, 289 300. https://doi.org/10.1190/1.1442303
475
Dahlin, T., 1996. 2D resistivity surveying for environmental and engineering
476
applications. First Break. https://doi.org/10.3997/1365-2397.1996014
477
Dahlin, T., Zhou, B., 2004. A numerical comparison of 2D resistivity imaging with 10
478
electrode arrays. Geophys. Prospect. 52, 379 398.
479
https://doi.org/10.1111/j.1365-2478.2004.00423.x
480
de Groot-Hedlin, C., Constable, S.C., 1990. Occam’s inversion to generate smooth,
481
two-dimensional models from magnetotelluric data. Geophysics 55, 1613 1624.
482
https://doi.org/10.1190/1.1442813
483
Fargier, Y., Antoine, R., Dore, L., Lopes, S.P., Fauchard, C., 2017. 3D assessment of
484
an underground mine pillar by combination of photogrammetric and geoelectric
485
methods. GEOPHYSICS 82, E143 E153.
https://doi.org/10.1190/geo2016-486
0274.1
487
Fikos, I., Vargemezis, G., Zlotnicki, J., Puertollano, J.R., Alanis, P.B., Pigtain, R.C.,
488
Villacorte, E.U., Malipot, G.A., Sasai, Y., 2012. Electrical resistivity tomography
489
study of Taal volcano hydrothermal system, Philippines. Bull. Volcanol. 74,
490
1821 1831. https://doi.org/10.1007/s00445-012-0638-5
491
Gresse, M., Vandemeulebrouck, J., Byrdina, S., Chiodini, G., Revil, A., Johnson,
492
T.C., Ricci, T., Vilardo, G., Mangiacapra, A., Lebourg, T., Grangeon, J., Bascou,
493
P., Metral, L., 2017. Three-Dimensional Electrical Resistivity Tomography of the
494
Solfatara Crater (Italy): Implication for the Multiphase Flow Structure of the
495
Shallow Hydrothermal System. J. Geophys. Res. Solid Earth 122, 8749 8768.
496
https://doi.org/10.1002/2017JB014389
497
Günther, T., Rücker, C., Spitzer, K., 2006. Three-dimensional modelling and
498
inversion of dc resistivity data incorporating topography - II. Inversion. Geophys.
499
J. Int. 166, 506 517. https://doi.org/10.1111/j.1365-246X.2006.03011.x
500
Herd, R.A., Edmonds, M., Bass, V.A., 2005. Catastrophic lava dome failure at
501
Soufrière Hills Volcano, Montserrat, 12–13 July 2003. J. Volcanol. Geotherm.
502
Res. 148, 234 252. https://doi.org/10.1016/j.jvolgeores.2005.05.003
503
Lesparre, N., Grychtol, B., Gibert, D., Komorowski, J.-C., Adler, A., 2014.
Cross-504
section electrical resistance tomography of La Soufriere of Guadeloupe lava
505
dome. Geophys. J. Int. 197, 1516 1526. https://doi.org/10.1093/gji/ggu104
506
Levenberg, K., 1944. A method for the solution of certain problems in least squares.
26 Q. Appl. Math. 2, 164 168.
508
Lines, L.R., Treitel, S., 1984. A review of least-squares inversion and its application
509
to geophysical problems. Geophys. Prospect. 32, 159 186.
510
https://doi.org/10.1111/j.1365-2478.1984.tb00726.x
511
Loke, M.H., 2012. Tutorial: 2-D and 3-D electrical imaging surveys. Geotomo
512
Software, Malaysia.
513
Loke, M.H., Chambers, J.E., Rucker, D.F., Kuras, O., Wilkinson, P.B., 2013. Recent
514
developments in the direct-current geoelectrical imaging method. J. Appl.
515
Geophys. 95, 135 156. https://doi.org/10.1016/j.jappgeo.2013.02.017
516
Marescot, L., Rigobert, S., Palma Lopes, S., Lagabrielle, R., Chapellier, D., 2006. A
517
general approach for DC apparent resistivity evaluation on arbitrarily shaped 3D
518
structures. J. Appl. Geophys. 60, 55 67.
519
https://doi.org/10.1016/j.jappgeo.2005.12.003
520
Marquardt, D.W., 1963. An Algorithm for Least-Squares Estimation of Nonlinear
521
Parameters. J. Soc. Ind. Appl. Math. 11, 431 441.
522
https://doi.org/10.2307/2098941
523
Miallier, D., Boivin, P., Deniel, C., Gourgaud, A., Lanos, P., Sforna, M., Pilleyre, T.,
524
2010. The ultimate summit eruption of Puy de Dome volcano (Chaine des Puys,
525
French Massif Central) about 10,700 years ago. Comptes Rendus Geosci. 342,
526
847 854. https://doi.org/10.1016/j.crte.2010.09.004
527
Nicollin, F., Gibert, D., Beauducel, F., Boudon, G., Komorowski, J.-C., 2006.
528
Electrical tomography of La Soufrière of Guadeloupe Volcano: Field
529
experiments, 1D inversion and qualitative interpretation. Earth Planet. Sci. Lett.
530
244, 709 724. https://doi.org/10.1016/j.epsl.2006.02.020
531
Park, S.K., Van, G.P., 1991. Inversion of pole-pole data for 3-D resistivity structure
532
beneath arrays of electrodes. Geophysics 56, 951 960.
533
https://doi.org/10.1190/1.1443128
534
Portal, A., Gailler, L.-S., Labazuy, P., Lénat, J.-F., 2016. Geophysical imaging of the
535
inner structure of a lava dome and its environment through gravimetry and
536
magnetism. J. Volcanol. Geotherm. Res. 320, 88 99.
537
https://doi.org/10.1016/j.jvolgeores.2016.04.012
538
Revil, A., Johnson, T.C., Finizola, A., 2010. Three-dimensional resistivity tomography
539
of Vulcan’s forge, Vulcano Island, southern Italy. Geophys. Res. Lett. 37,
n/a-540
n/a. https://doi.org/10.1029/2010GL043983
541
Robain, H., Bobachev, A., 2017. X2IPI‑: user manual.
542
Rosas-Carbajal, M., Komorowski, J.-C., Nicollin, F., Gibert, D., 2016. Volcano
543
electrical tomography unveils edifice collapse hazard linked to hydrothermal
544
system structure and dynamics. Sci. Rep. 6, 29899.
545
https://doi.org/10.1038/srep29899
546
Rücker, C., Günther, T., Spitzer, K., 2006. Three-dimensional modelling and
547
inversion of dc resistivity data incorporating topography - I. Modelling. Geophys.
548
J. Int. 166, 495 505. https://doi.org/10.1111/j.1365-246X.2006.03010.x
549
Sato, H., Fujii, T., Nakada, S., 1992. Crumbling of dacite dome lava and generation
27 of pyroclastic flows at Unzen volcano. Nature 360, 664 666.
551
https://doi.org/10.1038/360664a0
552
Soueid Ahmed, A., Revil, A., Byrdina, S., Coperey, A., Gailler, L., Grobbe, N.,
553
Viveiros, F., Silva, C., Jougnot, D., Ghorbani, A., Hogg, C., Kiyan, D., Rath, V.,
554
Heap, M.J., Grandis, H., Humaida, H., 2018. 3D electrical conductivity
555
tomography of volcanoes. J. Volcanol. Geotherm. Res. 356, 243 263.
556
https://doi.org/10.1016/j.jvolgeores.2018.03.017
557
Tarantola, A., 2005. Inverse Problem Theory, SIAM. ed.
558
Tikhonov, A., Arsenin, V., John, F., 1977. Solutions of ill-posed problems.
559
Wadge, G., Ryan, G.A., Calder, E.S., 2009. Clastic and core lava components of a
560
silicic lava dome. Geology 37, 551 554. https://doi.org/10.1130/G25747A.1
561
Zlotnicki, J., Boudon, G., Viodé, J.P., Delarue, J.F., Mille, A., Bruère, F., 1998.
562
Hydrothermal circulation beneath Mount Pelee inferred by self potential
563
surveying. Structural and tectonic implications. J. Volcanol. Geotherm. Res. 84,
564
73 91. https://doi.org/10.1016/S0377-0273(98)00030-4
